Ok, the bug also occurs in Beta 12.

I've also tried using 100% filter instead of transparency
(rgbt <0,0,0,1>, rgbt 1, rgbf 1, rgbf <0,0,0,1>), no
apparent change.
When changing some settings(rgbt, object/macro/blob)
the frequency and order of the four "bug-types" seem
to change, but still, its all just random.

Oh, but when not using a media around all particles,
they also disappear in bunches or one by one, not
only all-or-none.

So, since I gather that there are some problems occuring
with macros, I'll look into the dozens I'm using and check
if I can make a change there.

For now, it's getting too late for me.

But I'm confident that someone will:
-confirm
-point me to the actual bug (like "typical macro-bug" or such)
-and finally, some genius will find either a roundabout way,
 or a solution.

Generally, I just want to say that, though some people
do get annoyed by bugs, I know that this is beta-testing,
and I'm actually more proud of being able to find a bug,
so that I can contribute to the work of the actual developers.

I'm just doing my tiny work of giving them the chance to make a
superb raytracer, so I'm not complaining.

After all, the P in POVRay stands for Patience.
V is for Victory, but that always comes after Patience.

--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

On Sun, 17 Mar 2002 03:03:00 +0100, Tim Nikias <tim### [at] gmxde>
wrote:

>So, comments, suggestions and hints could be pretty useful.

Something to try:

max_trace_lever 479 // 2*(number of particles) + 1


Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG      e-mail : pet### [at] tagpovrayorg

Post a reply to this message

max_trace_level is set to 256, which is the
upper limit POV gives. But that's not the problem,
because in just a frame before, everything is fine,
and then suddenly, all media vanish (or some etc)


--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

Okay, I've found that the macros shouldn't leak
anywhere. As mentioned in some bug-report
with leaking macros, I've put the brackets
( ) around the returned parameters and checked
if there are possibilities for side-effects.

I've also found the objects aren't black, but just
eradicate everything in the line of sight, making
the background visible. When using background{rgb x}
the missing media are red.
When using a sky-sphere, the sky-spheres pigment
is visible.
When placing another object behind the media
particles, that one is also rendered invisible, but
that already happened with the mesh, just another
backup test.

I've posted a small image to p.b.i. so that you can see
one of the more curious effects:
The media is visible in front of the sky-sphere,
but renders the mesh invisible.


--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

I've experimented a bit with the objects in the
scene and found that blob's aren't shaded properly
sometimes. They miss reflections, and there is
only a visible difference between lit and unlit areas.
Specular is missing as well.

So, I guess it has nothing to do with media, but instead
some memory leak, caused by some line in my code.

I'm not using functions, only various macros,
include-file depth of 2 (including a file which includes
files) and I have no clue what the reason may be. I'll
be studying the known bug list for now and look with
what I'll come up.

--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

In case its of interest: It seems that the
bug doesn't appear on POV-Ray 3.1g.

--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

Just wanted to point out that the bug still occurs on
POV-Ray Beta 14, either compile, using my
1.4 GHZ Athlon, 512 MB SD-RAM, Win98

I'll keep an eye on this problem, and try to
pinpoint the exact error that occurs. I once
thought it might be a bug in media-calculations,
but that's not the case, as blobs sometimes don't
trace correctly (unshaded, sort of) as well.

If anyone could reply and tell me that the problem
is noted somewhere, or if someone can point me
to a related bug, I'd be happy, but I'll just keep
quiet on this one for now until I notice it being gone, or
until the final release is to come out, but the bug
still pops up...


--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

On Mon, 25 Mar 2002 09:41:31 +0100, Tim Nikias <tim### [at] gmxde>
wrote:

>blobs sometimes don't
>trace correctly (unshaded, sort of) as well.

Blobs with numerous small components? Maybe it's a precision thing,
blobs are represented by quite a high order equation so the solver
might just be choking in some cases.


Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG      e-mail : pet### [at] tagpovrayorg

Post a reply to this message

Naaah, I meant wrong lighting. Its like using a lower
Quality setting for tracing, or perhaps that
"cartoon-shading" type of effect. The colors aren't
lit and shaded properly, no reflections or highlights
(phong or specular). I'll recreate that a bit later and post
it on the binaries.images as an attachment to the other
three bug-pics I already posted on this matter.


--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message

Still present in Beta15.

--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html

Post a reply to this message